Every paper checked — nothing hidden

All papers

Every paper checked for brief #4 of 12–19 Oct 2026 — nothing is hidden. Brief papers first.

All17,313 In brief24 More (on-topic)3,333 Not relevant13,754 Not assessed87 Older352 Could not confirm93 Blogs with no paper266 All blogs284

Paper	Topic	Relevance
From Volpiano to Pixels: 30,000 Lines of OMR Ground Truth for Medieval Chant The paper introduces an optical music recognition dataset and pipeline linking medieval chant manuscript images to musical target tokens. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant OMR dataset paper, not general multimodal reasoning or VLMs.
From Volpiano to Pixels: 30,000 Lines of OMR Ground Truth for Medieval Chant The paper introduces an optical music recognition dataset and pipeline linking medieval chant manuscript images to musical target tokens. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on historical manuscript transcription, not industrial visual inspection.
EviStress: Evidence Stress Learning for Generalizable Face Forgery Detection with Textual Explanations The paper introduces a framework combining targeted patch perturbation with textual explanations to improve face forgery detection generalization. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on face forgery detection, not industrial manufacturing inspection.
Barycentric Guidance: Turning Foundation Image Editors into Continuous Affective Controllers The paper proposes a training-free guidance method enabling continuous valence-arousal emotional control in foundation image editors. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Studies affective image generation control, not deployment safety or robustness.
Barycentric Guidance: Turning Foundation Image Editors into Continuous Affective Controllers The paper proposes a training-free guidance method enabling continuous valence-arousal emotional control in foundation image editors. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on text-guided affective image editing, not multimodal understanding.
Barycentric Guidance: Turning Foundation Image Editors into Continuous Affective Controllers The paper proposes a training-free guidance method enabling continuous valence-arousal emotional control in foundation image editors. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Does not address model compression, pruning, or efficient inference.
Barycentric Guidance: Turning Foundation Image Editors into Continuous Affective Controllers The paper proposes a training-free guidance method enabling continuous valence-arousal emotional control in foundation image editors. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Proposes an image editing controller, not LLM evaluation benchmarks.
Barycentric Guidance: Turning Foundation Image Editors into Continuous Affective Controllers The paper proposes a training-free guidance method enabling continuous valence-arousal emotional control in foundation image editors. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on facial emotion image editing, not industrial inspection.
Mining Cross-Modal Perception Gaps in Autonomous Driving: When Camera and LiDAR Disagree The paper analyzes camera and LiDAR detection discrepancies across driving datasets to evaluate assumptions in multimodal fusion design. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Examines camera-LiDAR sensor fusion for autonomous vehicles.
Mining Cross-Modal Perception Gaps in Autonomous Driving: When Camera and LiDAR Disagree The paper analyzes camera and LiDAR detection discrepancies across driving datasets to evaluate assumptions in multimodal fusion design. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on autonomous driving perception, which is explicitly excluded.
ConDenseDrive: Knowledge-Dense, Token-Sparse Visual Encoder for End-to-End Autonomous Driving ConDenseDrive combines knowledge distillation and token pruning to create an efficient visual encoder for end-to-end autonomous driving. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on autonomous driving perception, explicitly excluded from this topic.
ConDenseDrive: Knowledge-Dense, Token-Sparse Visual Encoder for End-to-End Autonomous Driving ConDenseDrive combines knowledge distillation and token pruning to create an efficient visual encoder for end-to-end autonomous driving. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on visual encoders for autonomous driving planning.
FOCUS: A Subspecialty Foundation Model for Intraoperative Central Nervous System Pathology FOCUS is a specialized pathology foundation model pretrained on intraoperative frozen-section slides of central nervous system tissue. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on intraoperative digital pathology, not manufacturing defect detection.
FOCUS: A Subspecialty Foundation Model for Intraoperative Central Nervous System Pathology FOCUS is a specialized pathology foundation model pretrained on intraoperative frozen-section slides of central nervous system tissue. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on pathology image classification, not multimodal models.
UniMetric: Universal Metric Scale Recovery for Visual Geometry Foundation Models UniMetric provides a training-free framework that recovers metric scale in visual geometry foundation models using object size priors. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Studies metric scale recovery for 3D reconstruction, not manufacturing.
UniMetric: Universal Metric Scale Recovery for Visual Geometry Foundation Models UniMetric provides a training-free framework that recovers metric scale in visual geometry foundation models using object size priors. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Addresses 3D geometric scale recovery rather than multimodal reasoning.
DDVR-Net: Degradation-Aware Decoupling and Visibility-Guided Restoration for Historical Document Recognition DDVR-Net decouples glyph structures from physical damage and uses visibility-guided restoration for historical document OCR. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on historical document OCR, not manufacturing quality inspection.
UniLAP: A Unified Latent Action Policy for Cross-Embodiment Manipulation UniLAP introduces a continuous latent action policy trained via flow matching to enable robot manipulation without relying on future frames. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on robotic arm manipulation policies, not industrial inspection.
UniLAP: A Unified Latent Action Policy for Cross-Embodiment Manipulation UniLAP introduces a continuous latent action policy trained via flow matching to enable robot manipulation without relying on future frames. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on robotics action representation, not multimodal reasoning models.
Self-Supervised Learning of Semantically Consistent Visual Features via a Simple 3D Prior The paper presents a self-supervised approach using a simple 3D prior to learn semantically consistent visual features for complex objects. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Investigates unimodal self-supervised visual feature correspondence using 3D priors.
Self-Supervised Learning of Semantically Consistent Visual Features via a Simple 3D Prior The paper presents a self-supervised approach using a simple 3D prior to learn semantically consistent visual features for complex objects. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Addresses 3D semantic correspondence, not industrial quality control.
MaskSense: Confronting the Visual Exploration Trap in Masked Image Generation MaskSense addresses decoding exploration biases in masked generative models when aligning them with reinforcement learning. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on masked image generation, not industrial defect detection.
MaskSense: Confronting the Visual Exploration Trap in Masked Image Generation MaskSense addresses decoding exploration biases in masked generative models when aligning them with reinforcement learning. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Studies masked generative image models, not multimodal text-vision interaction.
DIME: Scaling Facial Representation Learning via Differential Masked Autoencoding DIME scales facial representation learning using a differential masked autoencoder that supervises differences between identity-matched image pairs. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on vision-only autoencoders for facial representation learning.
DIME: Scaling Facial Representation Learning via Differential Masked Autoencoding DIME scales facial representation learning using a differential masked autoencoder that supervises differences between identity-matched image pairs. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on facial representation learning, not industrial inspection.
PerturbBridge: Direct Phenotypic Supervision for Cellular Perturbation Generation PerturbBridge directly supervises perturbation-specific phenotypes during generative training to model cellular responses in images. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Studies biological cellular image generation, not multimodal vision-language models.
PerturbBridge: Direct Phenotypic Supervision for Cellular Perturbation Generation PerturbBridge directly supervises perturbation-specific phenotypes during generative training to model cellular responses in images. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on cellular microscopy phenotypes, not industrial quality control.
A unified three-branch deep learning framework for automated detection and counting of wildlife in aerial imagery The authors present a three-branch deep learning framework for automated aerial wildlife detection, strip delineation, and image matching. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Addresses wildlife detection in aerial imagery, not manufacturing inspection.
A unified three-branch deep learning framework for automated detection and counting of wildlife in aerial imagery The authors present a three-branch deep learning framework for automated aerial wildlife detection, strip delineation, and image matching. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on unimodal aerial computer vision, not multimodal interaction.
Rethinking LLM-Guided Visual Question Answering: Does Deeper Knowledge Integration Really Help? The study compares LLM guidance integration strategies for visual question answering, showing shallow late fusion outperforms deeper attention injection. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Studies visual question answering, not industrial defect detection.
Scaling Unified Multimodal Models to Long-Horizon Interleaved Generation The paper introduces a visual-memory framework enabling unified multimodal models to handle long-horizon interleaved text and image generation. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Studies multimodal interleaved image generation, not industrial visual inspection.
Contrastive learning for video reuse detection The paper uses contrastive learning with a Siamese network to extract key visual frames and detect video reuse. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses strictly on video-to-video retrieval, not multimodal text-image models.
Contrastive learning for video reuse detection The paper uses contrastive learning with a Siamese network to extract key visual frames and detect video reuse. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Covers video copyright reuse detection, not manufacturing anomaly detection.
EgoProbe: Benchmarking Long-Range and Exact Memory in Streaming Egocentric Videos EgoProbe benchmarks long-range retention and event-specific recall under interference for streaming egocentric video models. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Evaluates streaming egocentric video memory, not industrial visual inspection.
MARS: Systematic Multi-Level Acceleration for World Action Models MARS accelerates world action model inference across transformer layers, diffusion denoising, and policy queries without retraining. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Accelerates world action models for robotics, not multimodal understanding.
MARS: Systematic Multi-Level Acceleration for World Action Models MARS accelerates world action model inference across transformer layers, diffusion denoising, and policy queries without retraining. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on robotic closed-loop control, not industrial vision.
Multics: Generative Multi-Player World Model with Hierarchical Communication Multics provides a hierarchical communication framework for multi-player interactive video world models to reduce quadratic attention scaling costs. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on multi-player video generation, not manufacturing inspection.
Multics: Generative Multi-Player World Model with Hierarchical Communication Multics provides a hierarchical communication framework for multi-player interactive video world models to reduce quadratic attention scaling costs. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Addresses multi-player video generation rather than cross-modal reasoning.
DexNinja: Learning Robust Dexterous Cutting Policy with a Real-to-Sim-to-Real Data Engine DexNinja introduces a real-to-sim-to-real framework to train robotic policies for dexterous cutting of deformable objects. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on dexterous robot cutting, not industrial defect inspection.
DexNinja: Learning Robust Dexterous Cutting Policy with a Real-to-Sim-to-Real Data Engine DexNinja introduces a real-to-sim-to-real framework to train robotic policies for dexterous cutting of deformable objects. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Deals with robotic manipulation, not multimodal vision-language reasoning.
AReA: Adaptive Receptive-Field Attention for Efficient High-Resolution Diffusion Transformers AReA uses an adaptive, coarse-to-fine receptive field attention mechanism to enable efficient high-resolution generation in diffusion transformers. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on attention mechanisms in image diffusion models rather than multimodal reasoning.
AReA: Adaptive Receptive-Field Attention for Efficient High-Resolution Diffusion Transformers AReA uses an adaptive, coarse-to-fine receptive field attention mechanism to enable efficient high-resolution generation in diffusion transformers. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on diffusion transformers for image generation, not industrial inspection.
Beyond Sparse Captions: Aligning Slide-Level Text and Patch-Level Vision in Pathology LIMIT aligns slide-level pathology reports with patch-level vision features by freezing the vision encoder and training only the text encoder. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Deals with clinical pathology slides, not industrial defect detection.
When Left Is Right? Camera-Motion Understanding beyond Screen CameraVLM couples motion states and on-screen flow dynamics to improve camera-motion understanding in vision-language models. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Investigates camera motion understanding in VLMs, not industrial inspection.
HybridDualPathNet: Cross-Fusion Attention of EfficientNetV2 and Swin Transformer for Brain MRI Classification of Alzheimer's and Parkinson's Disease The authors introduce a hybrid architecture combining EfficientNetV2 and Swin Transformer to classify Alzheimer's and Parkinson's disease from brain MRI. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Addresses single-modality medical imaging, not multimodal language models.
HybridDualPathNet: Cross-Fusion Attention of EfficientNetV2 and Swin Transformer for Brain MRI Classification of Alzheimer's and Parkinson's Disease The authors introduce a hybrid architecture combining EfficientNetV2 and Swin Transformer to classify Alzheimer's and Parkinson's disease from brain MRI. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Studies medical diagnostic prediction rather than generative AI deployment safety.
HybridDualPathNet: Cross-Fusion Attention of EfficientNetV2 and Swin Transformer for Brain MRI Classification of Alzheimer's and Parkinson's Disease The authors introduce a hybrid architecture combining EfficientNetV2 and Swin Transformer to classify Alzheimer's and Parkinson's disease from brain MRI. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Proposes a standard hybrid vision backbone, not model compression or efficiency.
HybridDualPathNet: Cross-Fusion Attention of EfficientNetV2 and Swin Transformer for Brain MRI Classification of Alzheimer's and Parkinson's Disease The authors introduce a hybrid architecture combining EfficientNetV2 and Swin Transformer to classify Alzheimer's and Parkinson's disease from brain MRI. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on medical brain MRI scans, not manufacturing inspection.
HybridDualPathNet: Cross-Fusion Attention of EfficientNetV2 and Swin Transformer for Brain MRI Classification of Alzheimer's and Parkinson's Disease The authors introduce a hybrid architecture combining EfficientNetV2 and Swin Transformer to classify Alzheimer's and Parkinson's disease from brain MRI. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Medical image classification; cs.SE category appears misapplied.
HybridDualPathNet: Cross-Fusion Attention of EfficientNetV2 and Swin Transformer for Brain MRI Classification of Alzheimer's and Parkinson's Disease The authors introduce a hybrid architecture combining EfficientNetV2 and Swin Transformer to classify Alzheimer's and Parkinson's disease from brain MRI. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Evaluates medical classification models, not language model benchmarks.
HairFlow: Modeling 3D Hair Strands with Multi-Modal Conditioned Diffusion HairFlow generates 3D hair strands using a multimodal diffusion framework conditioned on coarse meshes and multi-view sketches. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Covers conditional 3D mesh generation rather than multimodal reasoning or VLMs.
HairFlow: Modeling 3D Hair Strands with Multi-Modal Conditioned Diffusion HairFlow generates 3D hair strands using a multimodal diffusion framework conditioned on coarse meshes and multi-view sketches. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on 3D hair modeling, not industrial defect detection.
WIRE: Weight Space Implicit Representation Encoding of Frozen Video LLMs WIRE reduces video LLM inference costs by moving video computations from input tokens into temporary low-rank weight updates. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on video language model architecture, not industrial inspection.
Brain2Face: A Study on the Relation of EEG Signals and Spatio-temporal Head Avatars The paper presents a dataset and baseline model to track spatio-temporal human head avatars directly from EEG brain signals. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Addresses EEG-driven digital avatars, not industrial computer vision.
Brain2Face: A Study on the Relation of EEG Signals and Spatio-temporal Head Avatars The paper presents a dataset and baseline model to track spatio-temporal human head avatars directly from EEG brain signals. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on EEG-to-avatar synthesis rather than multimodal foundation models.
TableVision: Recovering Verified Structure from Historical Tables TableVision recovers verified table structure from historical documents using dynamic programming and column sum checks. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Targets historical document table extraction rather than industrial manufacturing inspection.
Efficient Fréchet Distributional Decoder Alignment for Latent Generative Models Proposes a post-training decoder alignment method for latent generative models to resolve distribution mismatches between training and generation latents. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Addresses image generation latent decoders rather than multimodal foundation models.
Efficient Fréchet Distributional Decoder Alignment for Latent Generative Models Proposes a post-training decoder alignment method for latent generative models to resolve distribution mismatches between training and generation latents. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 06:18	Not relevant Focuses on generative image decoder optimization, not industrial inspection.
Efficient Fréchet Distributional Decoder Alignment for Latent Generative Models Proposes a post-training decoder alignment method for latent generative models to resolve distribution mismatches between training and generation latents. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Addresses generative model post-training alignment rather than LLM benchmarking.
Efficient Fréchet Distributional Decoder Alignment for Latent Generative Models Proposes a post-training decoder alignment method for latent generative models to resolve distribution mismatches between training and generation latents. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Studies distribution alignment in image generation, not deployment guardrails or safety.

6,481–6,540 of 13,754← Previous Next →